Deep brain stimulation-induced local evoked potentials outperform spectral features in spatial and clinical STN...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Neurological Surgery

Background:

• Subthalamic nucleus (STN) deep brain stimulation (DBS) is a key Parkinson's disease (PD) therapy.
• Optimizing STN DBS lead placement and stimulation programming remains a clinical challenge.
• Current methods use imaging and microelectrode recordings (MER), with programming relying on time-consuming trial-and-error.

Purpose of the Study:

• To compare DBS-induced local evoked potentials (DLEP) with traditional spectral features for STN mapping.
• To evaluate DLEP's accuracy in localizing the STN and predicting post-operative programming outcomes.
• To develop methods for artifact-free DLEP recovery and amplitude estimation at stimulating contacts.

Main Methods:

• Evaluated DLEP in 39 STN DBS leads from 31 PD subjects using single-pulse and high-frequency (HF) burst stimulation.
• Developed a novel artifact-removal technique for monopolar DLEP recovery and amplitude estimation.
• Compared DLEP and spectral features (beta, high-frequency oscillations [HFOs], aperiodic components) against imaging and MER localization and clinical outcomes.

Main Results:

• DLEP demonstrated high STN spatial consistency (100% for single-pulse, 84.62% for burst), outperforming beta (89.74%) and HFO (82.05%) spectral measures.
• DLEP correlated better with clinical outcomes (single-pulse ρ = -0.33, HF burst ρ = -0.26) than spectral measures (beta ρ = -0.25, HFO ρ = 0.05).
• Single-pulse, low-frequency stimulation proved sufficient for DLEP-based STN mapping.

Conclusions:

• DLEP offer superior STN spatial specificity and correlation with post-operative programming compared to spectral features.
• Developed artifact-removal and amplitude estimation methods facilitate clinical translation of DLEP.
• DLEP are a robust, clinically valuable marker for DBS targeting and programming, potentially reflecting STN excitability.